- user input gets non utf8 chars stripped all text input other then file paths.
- python has the same limitations, it will raise an error on non utf8 strings except for paths use unicode escape literals so its possible to deal with saving to these file paths from python.
- new string functions
BLI_utf8_invalid_byte(str, len) returns the first invalid utf8 byte or -1 on on success.
BLI_utf8_invalid_strip(str, len) strips non utf-8 chars.
the render engine assumes the RenderResult's rectf is not in linear color space when color management is disabled so the sequencer and opengl render need to follow this else it results in dark images.
also made drawing in camera view stick to the camera border (belated durian request),
useful for animation review without worrying about screensize moving the overlay about.
- raise an exception when python calls is_property_set(name) or is_property_hidden(name) and the property does not exist.
- added BLI_findstring_ptr(), which finds a named item in a listbase where that name is a pointer to a string.
- replaced inline for loops with calls to BLI_findstring_ptr() and IDP_GetPropertyFromGroup().
For !WIN32 systems the fix was in ED_fileselect_set_params
(basically adding the first folder in the sfile->folders_prev list)
For WIN32:
I talked with Nathan (Jesterking) and he agreed that the fix in path_util.c was required too. Without that BLI_path_abs was always making WIN32 paths ending with \ to end with \\
(e.g. C:\Blender\ --> C:\Blender\\)
And this was making the folder to fail ISDIR tests.
replace some long duplicated, ifdef'd if statements for image extension.
- new function: BLI_testextensie_array(), can take an array of extensions.
- define extension arrays: imb_ext_image, imb_ext_movie, imb_ext_sound - we could have more of these.
- removed amiga extensions iff and lbm
In this commit i've fixed crash when opening file browser. This crash was caused
by different _stat structures size in storage.c and filelist.c (because of force
setting __MSVCRT_VERSION__ in storage.c), other errors from report
were fixed earlier.
I've used _stati64 strcutrure and functions and tested on official
mingw-gcc 3.4.5, official mingw-gcc 4.4.0, tdm-mingw-gcc 4.4.1,
cross mingw-gcc 4.4.4 and cross mingw-gcc 4.2.1-sjlj -- everything was
ok in this configurations (except openexr and opencollada libraries which
requires sjlj compiler, so i was unable to build blender with that
libraries by official mingw-gcc 4.4.0)
Hopefully last fix for path issues for this release
* The config folder should also be searched for in the 'local' path for local installations
(This code was already there, but removed in revision 30440)
fix [#22920] Clicking folder icons doesn't work in File browser and append
* issue was that on windows stat requires the trailing slash removed, even with mingw.
(since we are after ahoy, putting this into extra #define, code cleanup will follow after release)
- Reverted own changes made to _stat function BLI_exists
- Use __MINGW32__ instead of FREE_WINDOWS in BLI_exists
- Removed recently added lib dependencies for mingw
- Use _stat64i32 instead of _stat in BLI_exist
Needed for correct compilation by mingw, no sense for msvc
(_stat is equivalent to _stat64i32 in Visual C++ 2005, and later i suppose)
- Added library msvcr90 for mingw to solve linking error to _stat64i32
* rename BLI_gethome to BLI_getDefaultDocumentFolder to better reflect how this function is used
* replaced BLI_gethome with getenv("HOME") on Linux and Mac where it retrieves location of bookmarks that are inserted as system bookmarks. BLI_gethome was a thin wrapper around these and in this case the user's home directory is what is actually wanted.
* fix for autosave location -> shouldn't use BLI_gethome anymore
* this frees BLI_gethome of having to emulate the local->user->system search path and can now be truly considered as 'home/default location for .blend files'
* removed setting the default G.sce from read_history, was out of context there.
* fix for creating user dir, leftover from previous commit.
jesterKing, please review -> if there are any issues I will fix or revert.
What happens is that blender looks for a directory "python" in the same
place as the executable for local installations, but that also means when
you have /usr/bin/blender it will look for /usr/bin/python, which is an
executable. Now it checks if it is actually a directory and not a file.
separate define for the user and system blender directory name,
on Linux the directories should be named /usr/share/blender and ~/.blender.
Platform maintainers should still check if that's ok.
For user config and data files, first check the 'local' location (where the executable is located), and only then the actual user locations (whatever the convention for the OSes; $HOME, %APPDATA%, etc).
Update after discussions on IRC:
* operating system specific path retrieval is moved back to GHOST, nothing blender specific here though
* cleaned up path functions a bit to remove #ifdefs
* removed Cocoa from blenlib again
TODO:
* Matt, Damien, please check and correct the functions for Cocoa and Carbon, could only put back existing code but needs adjustment
* finish GHOST_getBinaryDir - this should replace the BLI_where_am_i eventually as well as BLI_getInstallationPath on Windows and get_install_dir for the blenderplayer runtime
* It would probably be nice to define GHOST_getTempDir as well and move those out
* more cleanups...
NOTE:
Things are likely broken for macs
